---
title: Use in CI
description: Run dualmark verify in your pipeline to catch AEO regressions before they ship.
---

`dualmark verify` exits non-zero when a required check fails, making it a natural fit for any CI
pipeline. Add it to your pull-request workflow and block merges the moment conformance drops.

## GitHub Actions

<Steps>

<Step>

## Add the workflow file

```yaml title=".github/workflows/aeo.yml"
name: AEO Conformance

on:
  pull_request:
  push:
    branches: [main]

jobs:
  aeo:
    runs-on: ubuntu-latest
    steps:
      - name: Setup Bun
        uses: oven-sh/setup-bun@v2
        with:
          bun-version: latest
      - name: Verify AEO conformance
        run: bunx @dualmark/cli verify ${{ vars.STAGING_URL }} --timeout 15000
```

Replace `vars.STAGING_URL` with the URL of your staging deployment, or hard-code a page URL
directly (e.g. `https://staging.example.com/blog/your-post`).

</Step>

<Step>

## Parse results in downstream steps

Use `--json` to get a machine-readable report you can forward to a dashboard or annotation step:

```yaml title=".github/workflows/aeo.yml"
- name: Verify AEO conformance
  run: bunx @dualmark/cli verify https://staging.example.com/blog/your-post --json > aeo.json

- name: Upload report
  uses: actions/upload-artifact@v4
  with:
    name: aeo-report
    path: aeo.json
```

<Callout type="info">
  Exit codes are the same with `--json` -- non-zero on required-check failure. The JSON report can be
  consumed by dashboards, Slack bots, or custom annotation scripts.
</Callout>

</Step>

</Steps>

## GitLab CI

```yaml title=".gitlab-ci.yml"
aeo-conformance:
  image: oven/bun:latest
  script:
    - bunx @dualmark/cli verify https://staging.example.com/blog/your-post --timeout 15000
  only:
    - merge_requests
    - main
```

For JSON output in GitLab, pipe to a file and expose it as an artifact:

```yaml title=".gitlab-ci.yml"
aeo-conformance:
  image: oven/bun:latest
  script:
    - bunx @dualmark/cli verify https://staging.example.com/blog/your-post --json > aeo.json
  artifacts:
    paths:
      - aeo.json
    when: always
```

## Docker / self-hosted runners

If your runner has Docker available but not Bun, use the official Bun image directly:

```bash
docker run --rm oven/bun:latest \
  bunx @dualmark/cli verify https://example.com/blog/your-post
```

Or in a `docker-compose`-based pipeline:

```yaml
services:
  aeo:
    image: oven/bun:latest
    command: bunx @dualmark/cli verify https://example.com/blog/your-post
```

## GitHub Action (coming soon)

A first-class `dodopayments/dualmark-action` is on the roadmap -- it will let you drop AEO
conformance into any workflow with a single `uses:` line and no Bun setup required. Track progress
in [issue #17](https://github.com/dodopayments/dualmark/issues/17).

## Next steps

<Cards>
  <Card title="The verify CLI" href="/docs/conformance/cli">
    Full flag reference, exit codes, and programmatic API.
  </Card>
  <Card title="Scoring & levels" href="/docs/conformance/scoring">
    What each check is worth. Basic / Standard / Advanced thresholds explained.
  </Card>
</Cards>